SAN FRANCISCO—Nvidia will partner with Illumina to apply genomics and AI technologies to analyze and interpret multi-omic data in drug discovery, clinical research, and human health, the companies said in a collaboration of technology leaders announced early during the 43rd Annual J.P. Morgan Healthcare Conference.
Illumina will offer its DRAGEN analysis software on Nvidia’s accelerated computing platforms and make Nvidia’s BioNeMo tools available on the next-gen sequencing giant’s Illumina Connected Analytics (ICA) platform. By integrating their technologies, the companies aim to expand the reach of DRAGEN and Illumina multi-omics analysis globally to wherever users access Nvidia’s computing platform.
“Nvidia has had great success in building GPUs as an acceleration platform, and they have a wide distribution of that asset. They have GPUs everywhere driven by the AI demand that has risen. We’ll work together to make sure that we can also run DRAGEN on GPUs,” Rami Mehio, Illumina’s head of global software and informatics, told GEN Edge.
That work, Mehio said, has been in progress for a year and is very far along, but will take some additional time to complete.
“Our code is quite big, and to basically port it all in an efficient way on GPU takes time,” Mehio explained. “Both teams, Nvidia and us, have been working together on it. We do have prototypes running today, but they’re not production quality yet. We’re still adding functionality and removing bugs.”
“I would expect that we will have something certainly before the end of this year that’s available,” Mehio added.
Integrating NVIDIA with DRAGEN and Illumina Connected Analytics helps Nvidia at a time when it is working with several undisclosed countries as they establish national sovereign AI strategies.
“Really unique opportunity”
“They want to protect their population data, but they also want to create the conditions that will enable them to have a world-class health system, all at the same time. So, we have a really unique opportunity by bringing DRAGEN in the Illumina platform on our GPUs,” Kimberly Powell, Nvidia’s vice president of healthcare and life sciences, told GEN Edge. “We’re going to have the ability to really bring all of this amazing genomics capability into these sovereign AI efforts and allow them to have state-of-the-art genomics capabilities in these nations.”
Nvidia and Illumina have also agreed to collaborate on advancing multi-omic data analysis using Illumina Connected Analytics, in addition to developing new biology foundation models, with plans to incorporate Nvidia’s image processing and single-cell tertiary analysis tools onto the Illumina Connected Software.
The companies said their researchers will also work to integrate through Illumina Connected Analytics an array of Nvidia technologies ranging from Nvidia RAPIDS™ accelerated data science software included in the Nvidia AI Enterprise software platform; to the Nvidia BioNeMo™ generative AI platform for drug discovery, including generative AI models and fine-tuning capabilities for proprietary datasets; and the Nvidia MONAI framework for spatial cell imaging workflows.
“By bringing them into the [Illumina] Connected Analytics platform, we’re going to essentially be able to broaden and grow the genomics ecosystem substantially into places like the research community, into places like drug discovery, into hopefully every clinic on the planet, because we’re moving away from just reading the human genome into what kind of insights can we gather from this human genome,” Powell said.
“There’s a deep desire by more and more parts of the industry that would like to discover more about biology, about disease, to use these tools in the efforts of finding new medicines or coming up with better treatment strategies,” Powell elaborated. “The challenge is, that the whole human genome is not yet human readable. It has to be machine-readable. Then, you want to continue to gain insights from that data. We’re really at the point that we have to start applying machine learning and artificial intelligence in a lot more serious ways.”
Visiting Jensen Huang
Illumina’s Mehio said the collaboration began exactly one year ago during last year’s J.P. Morgan Healthcare Conference, when he joined Illumina’s newly appointed CEO Jacob Thaysen, PhD, in meeting Nvidia’s founder and CEO Jensen Huang at his house.
“That kicked off our initial work on porting DRAGEN, but also with an eye on these bio-foundational models that we want to explore together, which really hold the promise of the future,” Mehio recalled.
The companies pursued two tracks as they began to collaborate. One was how to leverage the availability of GPUs with DRAGEN. “And the second part we basically discussed is how Nvidia and the world are in the midst of a revolution in AI, and genomics is something that would benefit from it,” Mehio said.
“Our first foray into that is to basically bring Nvidia’s AI and compute capability into our data platform, Illumina Connected Analytics, and complement it with our own AI efforts and analysis that we’re doing to drive that new space of foundational models for biology and so on,” Mehio summarized. “So, two pieces to the collaboration. We’ve progressed far on the first one, and we’re in the initial steps on the second one.”
The multimodal AI models expected to result from the companies’ efforts can uncover insights and streamline processes to boost the capabilities of human experts, Illumina CEO Thaysen told Huang during an informal “Fireside Chat” hosted by the Nvidia CEO Monday evening at the Fairmont San Francisco hotel, about a hilly half mile north of where the J.P. Morgan conference was taking place.
“Combining other information, other modalities, other ‘omics,’” Thaysen observed, “is going to give us much deeper insight into biology. But while DNA was very difficult itself, when you then combine all the omics, it becomes exponentially more challenging. It’s getting so complicated that we do need huge computing power and AI to really understand and process it.”
The companies said their partnership was intended to further expand the global genomics market—projected at $37.94 billion last year by Precedence Research in a report released December 31—and help enable breakthroughs in identifying targets, developing drugs in clinical phases, and discovering biomarkers.
Continuing to push
“There are a lot of reasons to continue to push the ability to reduce the cost and increase the speed, to introduce new approaches like artificial intelligence, and to really expand the market,” Powell said.
Addressing the 43rd Annual J.P. Morgan Healthcare Conference on Monday, Powell offered additional details of the collaboration and potential customers for the combined offerings.
“We’re going to work together to build new applications in Illumina Connected Analytics, bringing genomics foundation models, bringing single cell applications that are essentially real-time at very, very large scale, and expand the genomic accessibility and utility into new markets like drug discovery,” Powell said.
“We still have so much to learn about the foundational building blocks of biology. It’s in the amazing data that Illumina is generating, and now we’re going to be working together to build those tools, build those capabilities, and expand the genomics market into the next generation, into generative AI genomics, and all the way to making it more accessible to the research community, to the biopharma and biotech community, and even expanding it into clinical and research-clinical applications,” Powell added.
Steven Barnard, PhD, Illumina’s CTO, told GEN Edge that the Nvidia collaboration meshes with his company’s three-pillar approach to its technology strategy going forward. Pillar one entails continuing to offer market-leading instrumentation and sequencing. Pillar two, expanding Illumina’s content offerings into multi-omics, as articulated by CEO Jacob Thaysen, PhD, in a recent GEN interview.
Pillar three, Barnard elaborated, “is the ability to be essentially the de facto insight and interpretation engine for all this biological data.”
“Besides the software that sits on top of the hardware, having the best tools, having the best technology to support that third pillar is extremely important,” added Barnard, who joined Illumina in 1998 as the company’s first scientist and fourth employee.
IQVIA, Mayo Clinic, and Arc Institute
The Illumina collaboration was one of four partnerships announced today by Nvidia. The other three are:
- IQVIA—The global provider of clinical research services, commercial insights, and healthcare intelligence for life sciences and healthcare companies plans to use the Nvidia AI Foundry service to build custom foundation models on its more than 64 petabytes of information, coupled with its deep domain expertise. IQVIA is also developing agentic AI solutions—outfitted with Nvidia NIM™ microservices and Nvidia Blueprints—designed to speed research, clinical development, and access to new treatments.
- Mayo Clinic—The world’s largest integrated, not-for-profit medical group practice joined Nvidia to say they will “massively” accelerate the development of next-generation pathology foundation models, as a “cornerstone” for future AI applications in drug discovery, as well as personalized diagnostics and treatments. Mayo Clinic plans to deploy newly available Nvidia DGX™ B200 systems based on the Nvidia Blackwell architecture, offering 1.4TB of GPU memory per system, and Nvidia’s MONAI healthcare imaging platform.
- Arc Institute—The Palo Alto, CA-based biology/machine learning research organization will partner with Nvidia to develop and share AI models and tools intended to advance biomedical discovery. Arc researchers and Nvidia engineers will scale up foundation models for biology designed to generalize across DNA, RNA, and protein drug modalities, plus advance applications for drug discovery and synthetic biology across complexity scales, diseases, and evolution research. Nvidia said it provided Arc with expertise in large-scale model development as well as the Nvidia BioNeMo platform running on Nvidia DGX™ Cloud, designed for easy-to-use, optimized training; and Nvidia NIM microservices and Blueprints.
